Fair, but as far as controlled experiments are possible in this kind of thing, it doesn't get much better than two kids with a shared birth parent and the same adoptive parents/environment.

Anyway, you're not wrong about the chaotic environment that is child rearing— the other side of this coin is sets of identical twins with differing sexualities, eg: https://kinseyinstitute.org/news-events/news/2019-07-26-twin...

No, it doesn't get much better than that scenario. But something as simple as which other kids are sat next to them in school can completely change their childhood. Observations like these are the best we have, and should not be ignored, but there are just so many variables that can never bet (ethically) controlled.

100% agree. I think because it’s hard enough to discern between what I would call first order (dna caused personality traits) and second order (immediate family and early childhood experiences) contributions to personality and what experiences a child had, that most people ignore third order, friends, bad luck, basically everything else. Then, you throw in luck ...I mean who knows why anyone does anything bad or good.

I’m not religious, but the saying often attributed to Christ saying, “forgive them, they know not what they do,” is probably totally, 100% accurate. I would add we don’t really know why they do what they do either. Christians want that to mean specifically about the crucifixion, but I believe whoever wrote that down was trying to say something more general about human experience.

I know from personal experience, just how much which kids you sit next to from 1st-5th grade, as they are often the same in the US, can influence your life in extreme ways.

All this says to me, we should probably go easier on everyone, even the “worst” of us...

Even within organisms starting with identical “first order” characteristics the opportunity for variation is fascinating. Here is a podcast on the topic I found illuminating, it is with the author of a recent book The Hidden Half

http://www.econtalk.org/michael-blastland-on-the-hidden-half...

> Fair, but as far as controlled experiments are possible in this kind of thing, it doesn't get much better than two kids

Actually, controlled experiments on this could be a lot better than N = 2, and the small-N problem dwarfs the positive features of the experiment here.

This is actually incorrect reasoning.

The amount of variation in the human condition would completely negate any conclusions from a sample size of two, however close the conditions.

It's also really easy to explain how one kid does well and the other doesn't, look at any life situation. One kid falls in with bad friends, one does not, one has a hobby that takes them somewhere, one does not, one gets a DUI the other does not, one has an amazing Hockey coach, the other does not, one takes a risk and wants to be a rock start, the other does not etc. etc..